Long term biogeochemical dynamics following a hurricane manipulation experiment in a humid tropical forest in Puerto Rico

Background/Question/Methods

Humid tropical forests are globally important sources of carbon dioxide (CO₂), methane (CH₄), and nitrous oxide (N₂O), and thus play a significant role in atmospheric greenhouse gas concentrations. These ecosystems are characterized by high litterfall productivity, which provides a key substrate for greenhouse gas production. Models predict that tropical regions will experience increased frequency and intensity of storm events with climate change. We used a factorial canopy trimming and litter manipulation experiment to explore the effects of litter deposition and canopy disturbance on greenhouse gas emissions from tropical forest soils. We sampled surface flux chambers (n = 5/plot) monthly from three replicate plots in each of four treatments over seven years (two years pre-treatment and five years post treatment) in a lower montane subtropical wet forest in Puerto Rico. Treatments included canopy trimming with litter deposited in situ (hurricane simulation), canopy trimming with litter removed (canopy disturbance alone), intact canopy with litter deposition in the understory (litter deposition alone), and an unmanipulated control. Our goals were to determine the impacts of disturbance on intra- and interannual patterns in litterfall and greenhouse gas fluxes, explore relationships among these variables, and estimate the magnitude and direction of feedbacks to global climate change.

Results/Conclusions

Pretreatment data showed consistent high magnitude seasonal fluctuations in litterfall (0.5 to 1.5 g C m^-2 d^-1) and soil CO₂ emissions (2.4 to 4.8 g C m^-2 d^-1) in this relatively aseasonal forest. Patterns appeared to be driven by small changes in light availability and temperature. Disturbance decoupled these patterns. There was little to no seasonality in litterfall for approximately one year following canopy disturbance, after which seasonal patterns, but not the magnitude of litterfall, recovered. In contrast, canopy disturbance did not alter the seasonal fluctuations of soil respiration, but did lower the magnitude of CO₂ emissions by 13-17 % over the first year. Litter deposition under an intact canopy enhanced litterfall and soil respiration significantly. Both canopy opening and litter deposition increased CH₄ emissions from soils (up to 5 µg CH₄-C cm^-2 h^-1). Canopy opening appeared to have a larger impact on N₂O emissions than litter deposition. Our results indicate that litterfall and soil respiration exhibit significant seasonality in a relatively aseasonal tropical forest, and that canopy disturbance decouples these C fluxes for one or more years. Canopy disturbance and/or associated litter deposition significantly increased the global warming potential of this ecosystem.